Related Experiment Videos
Recurrent excitation in neocortical circuits
R J Douglas1, C Koch, M Mahowald
1Institute of Neuroinformatics, University and Eidgenössiche Technische Hochschule, Zürich Switzerland.
Science (New York, N.Y.)
|August 18, 1995
Summary
Cortical neurons use excitatory feedback to amplify signals and maintain proportional firing rates, even with recurrent connections. This network conductance improves models of visual cortex processing.
Area of Science:
- Neuroscience
- Computational Neuroscience
- Systems Neuroscience
Background:
- Most cortical processing models use serial feedforward excitation.
- Mammalian cortical circuits predominantly feature excitatory feedback connections from neighboring neurons.
- Existing models do not fully capture the anatomical and physiological roles of feedback excitation.
Purpose of the Study:
- To investigate the role of excitatory feedback in cortical processing.
- To demonstrate how neuronal discharge remains proportional to stimulus strength despite recurrent excitation.
- To incorporate feedback principles into models of cortical direction and orientation selectivity.
Main Methods:
- Analysis of neuronal populations in cat visual cortex.
- Characterization of excitatory feedback as "network conductance".
- Development of computational models of cortical circuits.
Main Results:
- Excitatory feedback acts as a network conductance, amplifying feedforward input signals.
- Neuronal discharge is maintained proportional to stimulus strength, preventing runaway excitation.
- Incorporation of feedback principles enhances models of cortical direction and orientation selectivity.
Conclusions:
- Excitatory feedback is crucial for signal amplification and stable neuronal firing in the cortex.
- Network conductance provides a framework for understanding recurrent cortical processing.
- These principles are fundamental to the design of cortical architectures for sensory processing.